Fool me once...
In 2018, a research and analysis company named Cambridge Analytica illegitimately gained access millions of Facebook users’ data. Cambridge Analytica used the data to create psychological profiles of undecided voters, pushing them towards supporting presidential canditate Donald Trump (Confessore 2018). Although all parties involved were investigated, each continues to operate without major disruptions, and parties involved in similar work remain untroubled.
The event was not the first data violation at scale.
The Cambridge Analytica/Facebook event should have been data’s Watergate. In one fell swoop, data capitalists demonstrated the vice grip they have on society; democracy, social media and online platforms, and individual agency itself were modulated and violated with data. Instead, society at large shrugged, mumbled empty remarks, and resumed business as usual.
In this piece, I will discern what phenomena enabled this anticlimax, and sketch a theoretical framework that shifts data representations and reenables resistance against this kind of event and its antecedents.
“The threats of surveillance are real, and we’re not talking about them enough. Our response to all this creeping surveillance has been largely passive.”
(Schneier 2016, 10)
Data scandal–that is, the large scale misuse or mishandling of data, and the accompanying media saga–appears in a variety of forms, each dangerous in its own way: from the revelations of dragnet surveillance of an unwitting nation, to personally tailored advertisements targeting pregnant teenagers, to the leaking of comprehensive financial records of millions of individuals. The most troubling aspect of data scandal, however, is not necessarily the events being reported, but the non response they elicit. That is, not the absence of response, but the absence of any meaningful critique or resistance.
Consider the Cambridge Analytica/Facebook saga recounted in the preface. Browsing online discussion forums ranging from reddit yields:
A non response, from these examples, tends to display three sentiments in particular: ignorance, fatalism, and apathy. I will argue that each of these is enabled by unique constructions of what Beer (2018) calls the data imaginary: a site where “perceptions of data are shaped and also where entry points created for drawing data into social processes and prioritising their use in ordering and governing” (Beer 2018, 7). The data imaginary shapes the social, economic, political, and other imperatives, uses, and responses to data, subsequently affecting the potentialities of critique and resistance,
The non response represents a dangerous trend, symbolising the decline, or even death of critique and resistance against data. Without critique and resistance, any phenomenon can proceed unchecked with the potential to wreak havoc. To combat that end, I will unpack the non response here in order to implicate specific data imaginaries and pave a path towards resistance.
The very fact that the “scandal” in data scandal exists implies an imaginary that situates data in the shadows, a transient puppet master. Obfuscating and opacifying data practices and data power blinds the wider public (and disciplinary powers) to the misuse of their data and the mechanisms enabling it, disabling resistance through obscurity.
In the realm of this imaginary, data as a form of government is made invisible to facilitate a “machinery that ensures dissymmetry” of power relations (Foucault 1995, 202). “To scrutinise others while avoiding scrutiny oneself is one of the most important forms of power” (Pasquale 2016, 9); making hidden the data flows of the individual and the data power used to subjectify and make them transparent hinders the individual’s ability to “verify” their subjugation to that same data power (Foucault 1995, 201): Critique and resistance are quelled by virtue of data capitalists using these agnotological1 techniques to wedge doubt into the question of whether data practices and power exist at all, or at the very least, when and where they are present.
As a consequence of hidden data practices, it is difficult to delineate data power, and more critically, its limits. Data are immensely transgressive, and cannot be contained by the enclosures of ownership, location, security, or financial barriers to obtainment. They can thus be put to use in any imaginable situation by any entity with the resources to mine them and actualise them. The infinite “rhizomatic” recombinance of data flows (Haggerty and Ericson, n.d., 614) dissolve any attempts at containment or limiting the application of data, resulting in a functional absence of limit to data power, and making it vastly “polyvalent in its applications” (Foucault 1995, 205). This ‘runaway data’ thwarts resistance through its endless fluidity and accelerated pace: “we cannot anticipate the type of patterns that will emerge until we run the algorithms” (Andrejevic 2013, 36).
Pushing data practice into obscurity hides the specific modes of operation that underpin data practice; that is, not what practices are being performed with data, but how those practices are performed. Shrouding those processes blocks resistance outright: any attempts to subvert or disturb the processes would be a stab in the dark. Methods that purport to resist the few processes whose details have been made visible cannot ensure they actually do so:
“Anonymising software may shield us for a little while, but who knows whether trying to hide isn’t itself the ultimate red flag for watchful authorities?”
(Pasquale 2016, 9)
“Black boxing” does not only ward off public or industry attempts to resist data practice and power, but legal and political attempts, too. Disciplinary frameworks are subsequently playing a constant game of catch-up, with progress made by inches whenever data scandal erupts and the shadows are briefly lit (Pasquale 2016, 221–22).
In those moments of clarity where data practice and power are illuminated, obscurity is replaced by doubt. The nature of dataveillance and resultant data scandal often sees the very media platforms sharing the scandal being implicated in it. Though a separate issue to dataveillance and data scandal, this level of media modulation permits the Orwellian creation and destruction of news stories; data capitalists can “exploit the fog of war to throw up a series of often contradictory narratives” sowing enough doubt that data practice and power remain obscured and opaque (Andrejevic 2013, 17).
“We believe that modern technology platforms, such as Google, Facebook, Amazon, and Apple, are even more powerful than most people realize... And this makes the people who build, control, and use them powerful, too.”
(Schmidt and Cohen (2014), cited in Schneier (2016, 68))
Contrasting the hidden data imaginary, though not antithesising it, is the quasi-zealous imaginary of dataism. Equally opaque in its detailing of data practice and power, dataism instead uses this opacity to swell the omnipotence of data, ai, and the algorithm, encouraging engagement with, and submission to, data. “Tell us everything, Big Data croons. Don’t be shy. The more you tell us, the more we can help you” (Pasquale 2016, 7). By virtue of presenting data as the ultimate mediative power, the dataism imaginary rejects critique and resistance outright.
As the techno-utopianism in dataism seduces and proliferates, data capitalists begin to fence society in a digital enclosure, where “every action, interaction, and transaction generates information about itself” (Andrejevic 2009, 53). Trapped between remaining in modernity or renouncing data, individuals in the digital enclosure are forced into ‘prosumptive slavery’, generating data in order to simply participate in society (Andrejevic 2009, 53–54). Resistance is halted out of the sheer requirement to challenge immensely powerful structural forces.
The structures exerting those forces, too, become functionally as omnipotent and infallible as the data they accumulate and exploit. Data capitalists emerge as hegemons of power, mediating and modulating social processes at their will (Andrejevic 2013, 20–24). Through the same processes that create ignorance of data and prevent the development of adequate disciplinary and resistive frameworks, data capitalists are not held accountable for data scandal because their exact transgressions are unclear in the eyes of policy and the law. Failing to implicate them normalises the data scandal as “individuals become utterly accustomed to being seen” forcing individuals into acceptance and resignation to subjectification (Smith 2018, 10).
Extremism on the data imaginary spectrum, whether manifest as transcience or dataism, elicits the non response out of a lack of resistive potential. Engendering ignorance or fatalism on the public, respectively, such imaginaries stifle resistance in unique ways, but ultimately with the same result. Conversely, centerism on the imaginary spectrum dismisses resistance as outright unnecessary, discouraging critical engagement with the processes of data.
In the rare case they are punished, punishment comes in the form of a ludicrously large fine that is still paid off effortlessly. There are no lasting repurcussions, legal or otherwise: consider Cambridge Analytica themselves. The company was “dissolved”, but in reality continues to trade under a new name, without any issues.
In order to identify the potential for resistance, yet forego engaging in it, there is an acceptance of (though not resignation to) data scandal detriments, either through minimisation of the perceived harm of those detriments, or maximisation of the perceived benefits of continued use of the implicated platforms and services. This skewed ‘data calculus’ has faded of resistive desire. Both sides of the skew are brought by the data imaginary deemed the “data doxa” (Smith 2018, 6). Data doxa is an invocation and recontextualisation of Bourdieu’s doxa, describing how normalisation of data flows, practices, and power serve to “divert [the individual’s] attention from interrogating the power asymmetries and relations sturcturing the ‘digital enclosure’” (Smith 2018, 7). Under the doxic veil, data flows and data power are perceptively sterilised, their powers of mediation and modulation, to recontextualise Browne (2015)’s term, unvisible: wilfully invisible (Browne 2015, 63).
Data and the data double are disembodied and, under doxic sensibilities, a discharge of the everyday (Smith 2018, 9–10). Interplay between the two is intangible and unvisible, and their modulative capabilities on the biographical self are backgrounded (Smith 2018, p1–2). When data scandal erupts, it is not the biographical self that is in immediate harm; only through the data double can data power actuate onto the individual. Such dissonance between the data double and the self, exacerbated by the ignorance of data power described previously, severs the perceived causality between data power and the biographical self, masking the subjectification of the biographical self. This severance poses a cognitive roadblock in victimising those affected by data scandal. Because there is no biographical self being directly affected, only a disembodied set of data flows remains to empathise with and victimise. In contrast, a classical moral panic centers on an identifiable, victimised individual or group. There are a biography and body to empathise with; data scandal offers neither.
As a disembodied, unvisible presence, data are remarkably easy to output without realising, moreso when digital enclosures and prosumptive slavery are normalised. By contrast, prosumptive platforms or other data-generative services offer significantly more tangible returns, whether promising to connect the user with friends, offering a platform to video stardom, or simply a convenient way to send money (Smith 2018, 11). This exchange highlights a bizarre data valuation paradox, in which data are trivial at the individual level, and the services received in return are far more valuable, yet data are immensely powerful at the structural level, and the investment into the services being offered pale in comparative value. This weighs at the opposite end of our data calculus, glorifying the utilities gained through prosumption and neglecting the value of data as a commodity and power as it is freely given away.
Furthermore, we encounter the dilemma of negligible damage even to the data double. So often, data scandals occur in the form of leaks–proliferation of usually private data. In many of these cases, though identity theft and fraud could subsequently arise, they never eventuate due to the resources required to parse and mine big data sets. It makes you visible and vulnerable to potential damage.
In a classical moral panic, damage is not just potentiated, but actualised. When a child is murdered, children are not only perceived to be more likely to be murdered, but one already has been; there is a reference point, and a clear victim. Data scandal offers neither.
Even data scandal itself becomes an inevitable. Smith (2018)’s invocation of Pierre Bourdieu’s doxa describes how “habituation makes things seem normal quotation”. Data scandal is subject to similar habituation, subsequently becoming expected, and accepted.
Whether stifled or discouraging resistance, the data imaginary holds demonstrable power over the potential to counter data scandal. Disrupting that hold, then, necessitas the disruption of the data imaginary itself. Though the particulars and vagaries of data will doubtlessly be the subject of ongoing debate, the more extremist framings can be trimmed to uncover data fact within the imaginary. Grounding the definitions and powers of data into reality more clearly illustrate their impact and mediation of the data double, as well as their fallibilities.
Reversing apathy, alternatively, requires shattering the data doxa. This is difficult to achieve at the macro level; causal factors for data doxa are entrenched in modernity. At the level of the self, however, the data calculus may be skewed in the other direction by increasing perceived damages. Welding the data double to the biographical self permits tangible victimisation of data scandal subjects, forcing reevaluation of the calculus and tipping the scales against any perceived benefits to reignite resistive potential.
Data are not transient and obscure, nor didactic and omnipotent. Quelling these extremist thoughts effectively cripples the mechanisms that enable structural forces to inflict data scandal without consequence. Understanding the data fact that lies within the embellishments of the data imaginary creates transparency that, while not “not just an end in itself”, is an ”interim step on the road” that allows interrogation of data practices, power, and scandal (Pasquale 2016, 46).
Briefly, a definition of data and their practices will include how data originate, whom they belong to, and the data processes that actualise them into data power. I caution that a dataveillance-first definition–one that does not favour the individual–is likely more damaging than extremist data imaginaries. A more just approach would derive the definition from the simple fact that data have the ability to affect human lives and agency. Beginning from this lemma, a definition that understands the modulative power of data can be achieved, granting the individual appropriate power to prevent slavery to data.
Data fact must also discuss the semiotics of ‘artificial intelligence’ and ‘algorithm’. Deconstructing these terms into what they are and do as data processes strips their monolithic aura and exposes them and their vulnerabilities to strategies of resistance.
A hybridised self
“Our physical bodies are being shadowed by an increasingly comprehensive ‘data body’. However, this shadow body does more than follow us. It does also precede us.”
(Stalder (2009, 120), cited in Smith (2018, 4))
The most salient aspect of apathy in the non response is its origin at the hand of “un-victimisation”: the failure to appreciate the modulative power of the data double– which is directly affected by data scandal–over the biographical self. It is difficult to construct a victim from a disembodied phenomena. Replacing the data ‘other’ with the data self, however, dissolves the boundary between data flows and the self. Moreover, it actuates the damage of a data scandal onto the individual’s most critical sense of self, the biography. This gives data scandal fangs, so to speak, enabling the construction of a victim by embodying the affected data flows.
With a properly constructed victim to empathise with, the detriments of data scandal can now be weighed more accurately in the data calculus. With those detriments more plain to see, individuals are motivated to consider resistance, and engage critically data scandal.
This piece has utilised literature on varying representations of data to inform understandings of the lack of critical engagement with large-scale data misuse and mishandling. So-called data scandals are seen to perpetuate through the ignorance, fatalism, and apathy that data imaginary engenders onto the public.
Strategies for reenabling resistance have been outlined. Through the dispelling of extremist facets of the data imaginary to achieve a data fact, opacities in data processes and data power can be made transparent, facilitating critique and resistance.
The theorisation in this piece begs for empirical study; detailed ethnographies of individuals and their unique perceptions of the data imaginary and data scandal would provide the insight needed to tailor the hybridised self and data fact into a suitable framework for resistance that the public can adopt and utilise. Such work is imperative to actualising resistance and stemming the flow of data scandal that continues to encroach on fundamental human privacy and agency.
Andrejevic, Marc. 2009. “Privacy, Exploitation, and the Digital Enclosure.” Amsterdam Law Forum 1 (4): 47–62. http://amsterdamlawforum.org/article/view/94.
Andrejevic, Mark. 2013. Infoglut: How Too Much Information Is Changing the Way We Think and Know. Routledge.
Beer, David. 2018. The Data Gaze: Capitalism, Power and Perception. Sage.
Browne, Simone. 2015. Dark Matters: On the Surveillance of Blackness. Durham, NC: Duke University Press.
Confessore, Nicholas. 2018. “Cambridge Analytica and Facebook: The Scandal and the Fallout so Far.” The New York Times. The New York Times. https://www.nytimes.com/2018/04/04/us/politics/cambridge-analytica-scandal-fallout.html.
Foucault, Michel. 1995. Discipline and Punish: The Birth of the Prison. Translated by Alan Sheridan. 2nd ed. NY: Random House, Inc.
Haggerty, Kevin D., and Richard V. Ericson. n.d. “The Surveillant Assemblage.” The British Journal of Sociology 51 (4): 605–22. https://doi.org/10.1080/00071310020015280.
Pasquale, Frank. 2016. Black Box Society: The Secret Algorithms That Control Money and Information. Harvard University Press.
Schmidt, Eric, and Jared Cohen. 2014. The New Digital Age: Reshaping the Future of People, Nations and Business. John Murray.
Schneier, Bruce. 2016. Data and Goliath: The Hidden Battles to Collect Your Data and Control Your World. W.W. Norton & Company.
Smith, Gavin JD. 2018. “Data Doxa: The Affective Consequences of Data Practices.” Big Data & Society 5 (1): 2053951717751551. https://doi.org/10.1177/2053951717751551.
Stalder, Felix. 2009. “Privacy Is Not the Antidote to Surveillance.” Surveillance & Society 1 (September). https://doi.org/10.24908/ss.v1i1.3397.
From Pasquale (2016, 8): the “structural production of ignorance”↩